|
|
| |||||||||||||||||||||||||||||||
Background The inhalation of soybean dust released during the unloading of soybeans into a silo caused outbreaks of asthma in Barcelona, Spain. This study was designed to determine morbidity due to asthma and serum IgE reactivity before and after the installation of filters in the silo.
Methods We measured soybean-allergen concentrations in 136 samples of air collected for 9 months before and 24 months after the installation of filters. We compared the number of days on which there was an unexpected increase in emergency room visits for asthma, the number of days on which the number of emergency room visits for asthma in one four-hour period was so high that it was unlikely to be due to chance, and the mean daily number of emergency room and intensive care unit admissions for asthma for a total of 60 months before and after filter installation. Serum IgE antibodies against soybean allergens were measured in 38 patients before and after filter installation.
Results The concentration of airborne soybean allergens on days when soybeans were unloaded decreased from 324 to 25 U per cubic meter after the installation of filters (P<0.001). The number of days on which there was an unusually large number of visits to the hospital for asthma and the number of days on which asthma was epidemic both decreased significantly (both P<0.001), from 29 to 6 and from 18 to 0, respectively. The mean daily number of emergency room and intensive care unit admissions for asthma on days when soybeans were unloaded decreased from 8.3 to 5.4 and from 0.26 to 0.01, respectively (both P<0.001). The mean serum IgE antibody concentrations in the 38 patients studied decreased from 2 Pharmacia reference units per milliliter to 1 (P<0.001).
Conclusions Installing filters on silos to prevent airborne dissemination of allergenic soybean dust eliminates outbreaks of asthma caused by inhalation of the dust, thus supporting the idea that the avoidance of allergens helps prevent asthma.
Methods
Measurements of Airborne Soybean Allergens
Aerosol samples were collected on fiberglass filters for 24 hours with high-volume suction pumps (model CAV-P, MCV, Collbato, Barcelona) located in the urban area where most cases reported during asthma epidemics occurred. A single sample of air was obtained every eight days from January 1987 through September 7, 1987, when silo 1 was used without filters; from September 8, 1987, through December 1987, when unloadings at the silo were canceled; and from January 1988 through December 1989, after the installation of filters on silo 1. The sampling frequency was established before the cause of the outbreaks of asthma had been identified. Air samples were obtained on 136 days, including 34 days when ships were unloaded and 102 when they were not. The fiberglass filters were stored frozen at -20 °C until assay, at which time allergens were eluted from the filter sheets and assayed with a radioallergosorbent inhibition test as described by Agarwal et al11. Reference soybean allergens were prepared as previously described5.
Indicators of Asthma
The number of days on which there was an unusually large number of emergency room visits for asthma, the number of days on which asthma was considered epidemic, and the number of admissions for asthma to emergency rooms and intensive care units on days when soybeans were unloaded and on days they were not were determined from January 1985 through September 1989. These data were obtained from the emergency room records of the four largest urban hospitals, which accounted for 75 percent of emergency room admissions for asthma in the city12,13. Since few children were affected during the epidemics of asthma,12 only patients who were at least 15 years old were studied.
The number of cases of asthma was considered to be unusually large on a given day if the number of emergency room visits for asthma on such a day was so high that the probability that such a number or a higher one was the result of chance was 0.025 or less2. This level of probability was calculated by assuming a Poisson distribution, with the 15-day moving average representing the number of cases expected14. The number of cases of asthma was considered to have reached epidemic proportions on a given day if the number of emergency room visits for asthma in one four-hour period was so high that it was unlikely to be the result of chance (P 
0.05). These four-hour clusters were assessed by the Knox and Lancashire approximation15 of the scan method16.
Soybean Allergy
To assess the decrease in serum concentrations of IgE antibodies against soybean allergens after the installation of filters, we studied a representative sample of 148 patients who were seen at an emergency room on a day when asthma was considered to be epidemic after the installation of filters. These patients had previously been included in a case-control study, the methods and partial results of which were reported elsewhere17. In a study of patients seen during an asthma epidemic before the installation of filters,3 the prevalence of IgE reactivity against a commercial soybean extract was 74 percent. In addition, we measured IgE antibodies against soybean allergens in serum samples collected from 38 patients during the last epidemic before the filters were installed and two years after filter installation. There were no statistically significant differences in age, sex, place of residence, frequency of atopy, and smoking habits between these 38 patients and the random sample of 148 patients with symptomatic asthma during an asthma epidemic who were studied only after the installation of filters.
Specific IgE antibodies against the commercial soybean antigens were identified by paper radioimmunosorbent assays and radioallergosorbent assays (Pharmacia, Uppsala, Sweden). The results were expressed as a continuous variable with the categorical radioallergosorbent-test scale, in which the values are expressed as the number of Pharmacia reference units (PRUs) per milliliter3. Serum samples obtained before the installation of filters were stored at -20 °C and then assayed at the same time as the samples obtained after installation.
Cutaneous hypersensitivity to the soybean-allergen preparation was assessed by standard skin-prick tests in 115 patients two years after the installation of filters. The antigens used were obtained from samples of the husks and dust of soybeans unloaded in Barcelona harbor at the time of the previous asthma epidemic. A skin reaction was considered to be positive when the diameter of the resulting wheal was at least 3 mm.
Statistical Analysis
The concentrations of airborne soybean dust, indicators of asthma, and levels of serum IgE antibodies against soybean allergens before and after the installation of filters were compared by two-tailed paired and unpaired t-tests and Fisher's exact probability test. Concentrations of airborne soybean dust and serum IgE antibody concentrations against soybean antigens were normalized by logarithmic transformation. The influence of the proximity of the patient's residence to the soybean point source, a history of occupational exposure to soybeans, the ingestion of foods likely to contain soybeans, age, sex, and serum IgE concentrations on the rate of desensitization to soybean antigens was assessed by logistic-regression modeling18 (Egret statistical package, Statistics and Epidemiology Research, Seattle).
Results
The concentration of soybean allergens in the air in Barcelona decreased significantly after filters were installed in silo 1 (geometric mean concentrations before and after the installation of filters, 54 and 20 U per cubic meter, respectively; P<0.001). This reduction was due to a marked decrease in the concentration of soybean allergens on days when soybeans were unloaded (geometric mean of 324 U per cubic meter [range, 10 to 10,590] in 12 samples collected before the installation of filters and of 25 U per cubic meter [range, 5 to 165] in 22 samples collected after the installation of filters, P<0.001) (Figure 1). There was no significant difference in the concentration on days when soybeans were not unloaded (19 U per cubic meter [range, 9 to 1035] in 18 samples collected before installation, and 18 U per cubic meter [range, 7 to 426] in 84 samples collected after).
|
|
|
Of the 38 patients seen during an asthma epidemic for whom serum samples had been obtained before and after the installation of filters, 29 (76 percent) had measurable concentrations of IgE antibodies against soybean allergens two years after the installation of filters, as compared with 32 (84 percent) before the installation of filters. The mean serum concentration of IgE antibodies against soybean allergens in these 38 patients decreased from 2 PRU per milliliter before the installation of filters to 1 PRU per milliliter afterward (P<0.001). When the results of the radioallergosorbent test were expressed semiquantitatively, 18 patients had a decrease in specific IgE antibodies after the installation of filters, whereas 2 patients had an increase (P=0.006). The rate of the decrease in the serum concentrations of IgE antibodies against soybean allergens was not related to the proximity of the patient's residence to the soybean point source, a history of occupational exposure to soybeans, the ingestion of foods likely to contain soybeans, age, sex, or serum IgE concentrations.
Discussion
The installation of filters in a particular silo in Barcelona harbor was very effective in reducing the amount of soybean dust in the air and in preventing subsequent outbreaks of asthma. Not only did outbreaks no longer occur, but the number of days on which there was an unusually large number of cases of asthma also decreased. In previous studies by our group,2,3,4 epidemiologic criteria for a causal relation between the unloading of soybeans and epidemics of asthma were fulfilled, except for a dose-response relation, but evidence of the effect of filter installation was still preliminary6. The cessation of epidemics of asthma after the installation of filters provides strong confirmation of such a causal relation.
The results of this study also demonstrate the impact of the installation of filters on other indicators of asthma. Two aspects deserve special consideration. First, the most marked reduction in the mean number of emergency room admissions for asthma on days when soybeans were unloaded occurred among men over 45 years of age. According to our estimates, this group would have accounted for 48 percent of the emergency room admissions for asthma that would have occurred from October 1987 through December 1989 had filters not been installed in silo 1. This finding is consistent with that of a previous study,17 which demonstrated that older men with asthma in Barcelona were more susceptible to epidemics of asthma, and confirms one of the most striking features of asthma epidemics due to soybean dust: older subjects have the highest rates of asthma attacks.
Second, there was a substantial reduction in the number of admissions for asthma to intensive care units. The rate of admissions on days soybeans were unloaded dropped from 26 per 100 days to 1 per 100 days after the installation of filters, so that admissions for asthma to intensive care units have become rare in Barcelona (2.1 admissions for asthma per million inhabitants per year). In a previous study,19 we found that most of the critically ill patients hospitalized during days on which asthma was epidemic had an abruptly occurring asthma attack followed by rapid relief of symptoms after conventional treatment in the intensive care unit. A similar course of illness was reported by Wasserfallen et al.20 and O'Hollaren et al.,21 suggesting that the severe attacks are caused by massive exposure to an inhaled allergen. However, the causal evidence in these later studies was mostly circumstantial22. In contrast, we found a clear causal link between the inhalation of an aeroallergen -- soybean dust -- and the occurrence of life-threatening asthma. The results also demonstrate that the identification of the source of an allergen and the avoidance of exposure to it can play a substantial part in the prevention of severe, acute asthma. Before the installation of filters in Barcelona, at least 26 people died during the outbreaks with symptoms consistent with those of an asthma crisis.
We found that serum IgE antibodies against soybean allergens persisted for two years after the installation of filters and the cessation of epidemics of asthma. The actual duration of the persistence of serum IgE antibodies could be even longer if our analysis was biased by regression toward the mean. This possibility cannot be excluded, since the results were based on a single preintervention measure. This relatively high residual level of sensitization to soybean allergens two years after the installation of filters is consistent with the results of other studies on occupational23,24,25 and nonoccupational26,27 asthma. The persistence of serum IgE antibodies and skin reactivity to soybean allergens after a long period of avoidance of exposure could be a result of the retention of the allergen in the lung or extrapulmonary tissue or of the autonomous production of IgE by B cells, as has been suggested by Venables et al23. However, persistent exposure to the background levels of soybean allergen in the air in Barcelona after the installation of filters could also explain the continuing sensitization.
Although evidence derived from studies of asthma epidemics cannot be directly extrapolated to cases of asthma that occur at other times, we conclude that more attention should be paid to the primary prevention of asthma through the precise identification of causative allergens and the subsequent avoidance of exposure to these allergens.
Supported by a grant from the Fondo de Investigacion Sanitaria (FIS-89/ 924-3).
We are indebted to Carlos Murillo, Ph.D., Prof. A.J. Newman Taylor, and Francisco X. Real, M.D., for their valuable scientific advice; to the members of the Barcelona Asthma Collaborative Group, who participated in various aspects of this study; to Ms. Helena Martinez for assistance in the preparation of the manuscript; and to Marta Pulido, M.D., for editing the manuscript.
Source Information
From the Departament d'Epidemiologia i Salut Publica, Institut Municipal d'Investigacio Medica, Barcelona, Spain (J.M.A., J. Sunyer, J. Sabria, F.M., M.S.); Allergic Diseases Research Laboratory, Mayo Clinic and Mayo Foundation, Rochester, Minn. (C.E.R.); Servei de Bioquimica (R.C., M.J.R.) and Servei de Pneumologia (F.M.), Hospital Vall d'Hebron, Barcelona, Spain; and Servei de Pneumologia i Allergia Respiratoria (R.R.-R., J.R.), Hospital Clinic i Provincial, Barcelona, Spain.
Address reprint requests to Dr. Anto at the Departament d'Epidemiologia i Salut Publica, Institut Municipal d'Investigacio Medica, Doctor Aiguader 80, 08003 Barcelona, Spain.
References
| |||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
Previous
